A. Introduction: Remembering Neelan Tiruchelvam
I am very honored to give this year’s Neelan Tiruchelvam Memorial Lecture. I had never had the privilege of meeting Neelan, to my deep and lasting regret. Neelan was the kind of constitutional lawyer that we all aspire to be. He was a voice of principle, clarity and common sense in the tumultuous debates over Sri Lanka’s constitutional future. Like many Sri Lankans of his generation, Neelan was deeply involved in the practical world of constitutional design and interpretation. Yet Neelan stood out because he anchored his views on legal issues in a deeper conceptual framework that was both interdisciplinary and comparative. For Neelan, Sri Lanka’s constitutional politics were very much a reflection of broader political debates over the fundamental character of the Sri Lankan polity, and were rooted in this country’s colonial history, post-independence politics, economic development, and linguistic make up. But Neelan also believed the constitutional issues that have confronted Sri Lankans since independence were not peculiar to Sri Lanka. The Sri Lankan dilemma was not just a Sri Lankan dilemma. There was much to be gained from engaging with the experiences of other countries facing similar challenges. I hope to do justice to his memory today by bringing global experience to bear upon Sri Lanka’s current predicament.
How Neelan approached Sri Lankan constitutional debates says something important about the politics of constitutional design in plural societies. Let me begin by posing a question: what is a plural society? A plural society is not merely a society which is ethnically, linguistically, religiously, or culturally diverse. What marks a plural society is that these differences are politically salient – they are persistent markers of political identity and bases for political mobilization. Ethnocultural diversity translates into political fragmentation. In a plural society, political claims are refracted through the lens of ethnic identity, and political conflict is synonymous with conflict among ethnic groups.
A lot is at stake in how plural societies respond to the challenges raised by the equation of ethnocultural identity with political interest. The extreme consequences of the failure to address these challenges are well known: discrimination and exclusion, forced assimilation, civil war, ethnic cleansing, and even genocide. But even in the absence of violence, failing to respond to these challenges appropriately can have a corrosive effect on ordinary politics. In the absence of trust and expectations of reciprocity across ethnic lines, it may become impossible to reach political decisions on routine questions of public policy. If the burdens and benefits of those policies are perceived to be unevenly distributed between ethnic groups, they may be condemned as discriminatory. Political debates on ordinary questions of public policy can escalate quickly into political dramas of respect and recognition far removed from the actual interests at play.
A constitution in a plural society bears a particularly heavy burden, because it plays multiple roles. I want to distinguish between two accounts of the function of a constitution in a plural society – the regulative conception and the constitutive conception.
On the regulative conception, constitutions both enable and disable political decision-making. They enable decision-making by creating the institutions of government, by allocating powers to them, by setting out rules of procedure to enable these institutions to make decisions, and by defining how those institutions interact. Constitutions also disable decision-making, by enacting procedural roadblocks (such assupermajority rules) and setting substantive limits on political decision-making (such as bills of rights).
In a plural society — as in any society — a constitution must fill this regulative role. But in a plural society, a constitution must go further and constitute the very demos which governs itself through the constitutional regime. Because of a history of conflict, or a lack of a shared existence, the constitution is often the principal vehicle for the forging of a common political identity, which is necessary to make that constitutional regime work. The constitution can foster the creation of a common political identity by creating the institutional spaces for shared decision-making amongmembersofdifferentethnicgroups. Concreteexperiencesofshared decision-making within a framework of the rule of law, without recourse to force or fraud, can serve as the germ of a nascent sense of political community. The process of debating and negotiating a constitution, if conceptualized as an inclusive process, can also help to create the political community on whose existence the constitutional order which results from that process depends. A constitution can also constitute the demos by encoding and projecting a certain vision of political community with a view to altering the very self-understanding of citizens.
Constitutional politics in plural societies takes place on two levels. On the one hand, there is a technical politics about the wording of constitutional texts, the design of basic institutions, and the interpretation of constitutional provisions. But contending positions on a wide number of specific and narrow questions draw on deeper conceptual models which themselves grow out of the practical experience of trying to fashion stable and legitimate frameworks for constitutional government in plural societies. Disagreements on particulars often reflect disagreements on fundamentals. Getting a handle on what these conceptual frameworks are is therefore practically important.
This evening, my principal task is set out two of the major conceptual frameworks for constitutional design in plural societies, which have been termed integration and accommodation. The debate between these models has become as polarized as ethnic conflict itself. These two models are currently at play in Sri Lankan constitutional debates over how to fashion Sri Lanka’s post war constitutional settlement. After setting out what those models mean in practice for constitutional design, I want to suggest how we should move beyond the sharp dichotomy between integration and accommodation.
B. Integration
Integrationism is an umbrella category. But in the contemporary scholarship and advocacy on constitutional design for plural societies, it proceeds from two premises.
The first premise concerns the constitutional significance of plural identities. For integrationists, ethnocultural diversity is a reality that cannot be denied. However, they relegate these identities to the private sphere, and do not give them any significance in the design of public institutions. The model here is the treatment of religion in liberal democracies. In order to prevent religious diversity from translating into political division, liberal democracies set out two constitutional principles, non-endorsement and non-interference. Non-endorsement means that the state is religiously neutral, and has no official religious identity. Non- interference means that religious identity is a strictly private matter, with individuals free to choose their religious identities free from force or fraud within a framework provided by the rule of law. Integrationists would extend this approach from religion to ethnicity race, and culture. The goal is to privatize these identities, and to constitute a public sphere around a common legal status enjoyed by all individuals, to which ethnicity is irrelevant. This is the civic conception of citizenship. On its purest form, civic citizenship is highly abstract, and is built around the shared principles of liberal political morality, such as respect for basic human rights, and democratic governments elected at regular intervals on the basis of universal suffrage.
The second premise concerns the character of political competition, and emerges from pluralist accounts of democratic politics. The question posed by pluralists is why political actors who lose within democratic institutions do not respond to those losses by turning on the system itself and attempting to undermine it. The answer is the theory of crosscutting cleavages, which holds that individuals belong to a number of different interests and outlooks. Crosscutting cleavages have two moderating effects. First, because individuals are members in multiple social groups, they will come into contact with a multiplicity of perspectives, and will possess a complex set of interests, which will tend to moderate their political attitudes. Second, in the absence of sharp partisan division among individuals, political elites will face pressure to moderate their political positions. This account of the nature of political cleavages is closely tied to the case for a competitive model of democratic politics. On the pluralist view, politics is characterized by shifting coalitions and majorities, which change from issue to issue, and do not endure over time. Political parties compete for median voters at the center of the political spectrum, which promotes moderation. The assumption is that parties will cycle in and out of government, as they assemble shifting coalitions of voters in their competition for the political center. Since there is no permanent exclusion of any segment of society from political power, the loser accepts this loss in the hope they will win another day.
These twin assumptions generate a series of concrete proposals of what a constitution should look like:
• It should contain a bill of rights that guarantees a universal set of liberties and freedoms on all citizens on a basis of equality, backed up by judicial review. Overtime, the standard package of what rights are contained in a bill of rights has expanded, to encompass not only civil and political rights, but also economic and social rights. A bill of rights serves two functions. First, it encodes a vision of political community built around citizens who are equal bearers of constitutional rights
– a civic citizenship – whose political membership is unmediated by group identity. Second, it serves to check well-known pathologies in the legislative process that may lead to the enactment of laws that infringe basic rights even by a legislature that is committed to the idea of rights. These include, inter alia, insufficient information regarding the impact of a law on rights because of poor legislative fact-finding, a rushed legislative process triggered by panic at times of war or emergency, and the cognitive bias which results when a legislature does not contain representatives from those groups who are affected by a law and therefore is unaware of its potential consequences.
• It should create a constitutional framework for strong majority governments that can achieve policy coherence, and present a clear choice to the electorate and be held accountable for its decisions. This overarching goal is achieved through a number of constitutional mechanisms.
o With respect to executive-legislative relations, it suggests Parliamentary democracy, in which thegoverning party commands the confidence of the majority of the legislature, held in check by an official opposition. Cabinets are formed on a winner- take-all basis, with the opposition party focused on providing parliamentary opposition from outside the government, and with no right to cabinet representation. Legislative voting is by simple majority, with no special role for minority parties in the legislative process.
o The electoral system should be one that is designed to produce stable legislative majorities that have the ability to govern. Single member plurality voting (first past the post), which systematically produces disproportionality between the percentage of votes cast and seats awarded, fares best on this score. Conversely, integrationists disfavor proportional representation, because it rarely produces majority governments led by a single party. The electoral system should also discourage political mobilization on the basis of ethnicity, and conversely, the mobilization around non-ethnic interests. This also argues in favour of single member plurality and a common electoral roll, because it is based on geographic constituencies. If we assume that electoral districts are ethnically diverse, parties will have to base their appeals on non-ethnic criteria to secure electoral support. Integrationists fear that proportional representation opens the door to the rise of ethnic parties that can collect votes scattered across many separate areas.
The political party system should be aligned with the goals of the electoral system. Integrationists favour umbrella, brokerage parties that compete across the entire country, and which attract support from a variety of different economic interests and ethnic groups, and attempt to mediate among them within the party caucus. Single plurality voting creates the incentive for the creation for such parties. In addition, political party laws may use national presence requirements, ethnic party bans, and substantive restrictions on party platforms (for example, a prohibition on the advocacy of federalism) to block the creation of ethnic parties or erect significant roadblocks toward their establishment or operation.
o Integrationists favour a single official language, for use across all spheres of government activity: legislative debates, judicial proceedings, the internal workings of the civil service, the educational system, and public services. Moreover, integrationists assume thatthe government’s choice of official language will have the effect of setting the language of the economy. Integrationists justify the promotion of a common language for all political and economic intercourse to serve a number of goals. They argue that a common language is the precondition to mass democratic deliberation that transcends ethnic differences – a common language enhances social mobility by ensuring that all citizens are able to take advantage of public and private sector employment opportunities, a common language supports the development of national labour markets because it creates a mobile, national labour pool capable of working in any part of the country, and that a common language is part and parcel of the process of administrative consolidation and the growth of the modern state because it makes it possible for states to communicate directly with citizens and for a growing civil service to communicate internally.
But a point of genuine disagreement among integrationists is devolution or federalism:
o On the one hand, there are integrationists who oppose federalism on the basis of a constitutional theory that considers the entire territory of the state as belonging on equal terms to all of its citizens. The national territory is indivisible. It follows that any internal political divisions that allow a portion of the national political community to govern itself, even in limited policy areas, and which constitutionally protects these policy decisions against the will of the national majority is a violation of popular sovereignty. There is an equation of external sovereignty against foreign states, and internal constitutional sovereignty.
o On the other hand, there are integrationists who advocate
federalism because of the values it promotes, such as the enhancement of democratic self-government by multiplying the opportunities for public office-holding and democratic participation, the promotion of policy experimentation by allowing different sub-national jurisdictions to engage in policy innovation, the satisfaction of citizen preferences by enabling sub-national jurisdictions to offer different packages of public services and to set different levels of taxation, and the closely related argument of fostering inter-jurisdictional competition to enhance personal liberty and preserve open markets by empowering mobile citizens and capital. But these integrationists are opposed to federalism as a device that politically empowers ethnic groups. They adamantly oppose federalism where sub- unit boundaries are designed to turn an ethnic minority into a sub-national majority. For the same reason, integrationists would ensure that a national bill of rights acts as a strong check on sub- national autonomy.
Let me conclude this portion of my lecture with a caveatfortheintegrationist constitutional model I have set out. There are many countries whose constitutional culture is clearly integrationist which depart one or more respects from this model: o For example, the United States, where I currently live, is a core case of an integrationist country. But it has a Presidential system of government. Many countries in Latin America likewise have Presidential systems, but integrationist constitutional cultures. The same can be said of countries with semi – Presidential systems, most notably France, which is an exemplar of an integrationist country.
o With respect to electoral systems, European Parliamentary democracies uniformly opt for proportional representation, which has led to a long-standing practice of coalition government. But their constitutional cultures are nonetheless resolutely integrationist.
My answer is that the constitutional model I have set out is just that – a model. It represents an account of how to best implement the premises of an integrationist constitutional culture. There may be genuine debate over what constitutional arrangements best realize integrationist goals. Moreover, there can be compromises in how far the goals of the model can be realized in practice, because of competing political considerations. But a model provides some kind of starting point for political discussion.
C. Accommodation
Now let me turn to accommodation, which proceeds from twin premises that are direct responses to the foundational premises of integrationists.
First, integrationists claim that civic political identities are culturally neutral and that can serve as a common point of reference for members of different ethnic groups. But accommodationists counter that even in states that adhere to the integrationist credo, civic identities often privilege a dominant ethnic group. The principal source of evidence for this charge is a large body of historical and sociological research on the manner in which states have created overarching civic identities within specific national contexts. This practice is known as “nation-building,” and consists of public policies that promote a common language, a shared history and a shared culture, usually facilitated by the centralization of legal and political power. Nation-building emerged as a tool of political consolidation in Western Europe, and spread with the rise of nationalism to Eastern and Central Europe, and with decolonization to Asia and Africa. The theory behind nation-building is that citizens must identify with political institutions in order for them to agree to work within them and accept their decisions. A shared civic identity built around abstract principles of liberal political morality will not be enough to generate this kind of bond. Those principles need to be rooted in a story of the national political identity. But accommodationists argue that nation- building can operate as a vehicle whereby a politically powerful ethnic group universalizes its particular identity. Perhaps the leading example is language, a common element in nation-building. Official language status operates to distribute economic and political power. If a group’s language becomes an official language, it has an immediate advantage in access to public sector employment, political office, higher education, and public services. Indeed, the accommodationist critique of official monolingualism sounds in the same register as the justification of those policies -a single official language can serve to impede democratic participation, social mobility, and administrative efficiency.
Second, plural societies often witness a breakdown in political competition that converts it from a mechanism to moderate ethnic conflict into a force that makes it far worse. The competitive paradigm of democratic politics depends on two assumptions — that opposition parties will eventually share power and that, because of the shifting nature of majority coalitions, governing parties will not abuse their power. But these assumptions do not hold in plural societies, because political mobilization occurs on the basis of ethnic identity, and political parties respond by organizing themselves on this basis. Ethnic political parties do not compete for median voters, and indeed, do not compete across ethnic divides. The precise consequences of the institutions of majoritarian democracy in plural societies will depend on the precise demography of the polity in question. If there is a clear ethnic majority, the result is not a temporary minority that will eventually cycle into power, but a persistent minority that will permanently be in opposition and excluded from political office, and a permanent majority that will be free from the restraint that a future loss at the ballot box imposes on the abuse of power. Where there is no clear majority, coalitions of minorities can produce the same effect. Minorities that are persistent losers with no prospect of wielding power may sit outside of politics and turn on the system itself, potentially through violence.
While accommodationists agree on the rejection of integrationism, there is a sharp disagreement on how constitutions should be designed in response. The two major approaches are consociationalism and centripetalism.
The leading exemplar of consociational theory is Arend Lijphart. It has the following main elements:
• The consociational model is based on Parliamentary democracy for executive-legislative relations. Cabinets are power-sharing coalitions that include representatives of the major ethic groups, each represented by ethnic political parties, with proportionate representation. The pluralist conception of government and opposition, with parties cycling in and out of power, is gone.
• Moreover, the logic of proportionality is carried into the civil service, police, and military, where real executive power is wielded. In addition, political decision-making provides for group vetoes over matters of vital interest (e.g. education and language), because even with proportionate representation, members of ethnic minorities may be outvoted in cabinet or in the legislature.
• The electoral system should create the incentives for parties to agree on grand coalition cabinets with proportionate representation. Proportional representation is the voting system that does this best for two reasons. First, it allows for the legislative representation of territorially dispersed minorities by ethnic political parties who may be outvoted in single member districts. Second, it creates a fragmented legislature in which a single party is far less likely to hold an absolute majority. Leaders of ethnic minorities can leverage their legislative power to secure executive power-sharing through cabinet membership.
• Another feature of the consociational model is group autonomy, which can take a territorial or a non-territorial form. Territorial group autonomy entails devolution or federalism, but in a form that is specifically designed to empower territorially concentrated ethnic groups. Sub-unit borders would be drawn to ensure that ethnic minorities constitute sub-national majorities, and sub-units would have jurisdiction over subject-matters integral to a community’s cultural survival, such as education and official language policy at the provincial level. Non-territorial forms of group autonomy are also possible when populations are interspersed, for example, with respect to education, social services, and religious personal law. These forms of group autonomy can be combined. In addition, territorial group autonomy can occur at the level of municipal governments.
Consociational theorists have devoted surprisingly little attention to official language policy. But we can infer that they would favour official multilingualism, with the state operating in multiple official languages to the greatest extent possible. But it is important to note that the range of linguistic choice will vary depending on the institutional setting. For example, it is possible for the state to provide public services in many different languages. Indeed, the deployment of government personnel from an ethnic majority to areas populated by ethnic minorities, and their refusal or unwillingness to deliver public services in the language of the local population, has given rise to accusations of internal colonialism in many countries across the world. But forthe internal working language of government, there are strong practical reasons to converge on a single language, to permit efficient internal communication among civil servants. The broader lesson is that official multilingualism in one institutional setting does not necessarily connote the same degree of official multilingualism in other contexts.
• Finally, consociational theorists have had relatively little to say about bills of rights. But what often lies at the root of ethnic political mobilization is discrimination and exclusion, and the unequal enjoyment of basic liberties and freedoms. Guaranteeing fundamental human right son a fully equal basis should logically be part of the consociational constitutional package. This sets the stage for difficult internal conflicts between consociational arrangements that institutionalize and empower group identity and bills of rights that guarantee individual rights. I will say more about this later.
Centripetalism emerged as a critique of the consociational model, and was developed by Donald Horowitz. His major criticism of consociationalism is that it does not explain the incentive for leaders of a majority ethnic group to enter into a power-sharing arrangement with minority leaders if they can control the state without sharing power. Similarly, in situations where there is no dominant ethnic group, all that may occur is a coalition that includes enough ethnic groups to command a majority of the legislature. Legislative elections under proportional representation would only compound the problem. When political parties are organized on an ethnic basis, political competition crosses within ethnic groups, not across the ethnic divide. Ethnic parties that attempt to adopt moderate policies will come under attack from extremist parties in a process of ethnic outbidding. Ethnic parties will respond by shifting to the extremes. This imposes electoral penalties on interethnic coalitions, which makes them unstable.
Centripetalism offers a different constitutional package to respond to these concerns:
• Centripetalists advocate electoral systems that reward ethnic parties electorally that appeal across ethnic lines. Cross-ethnic support should offset electoral losses from ethnic competition on the extremes. The favoured electoral system is the alternative vote, whereby winning candidates must secure a majority of the votes cast in an electoral district. Voters rank candidates in order of preference, and if no candidate is successful after first preferences have been counted, the bottom candidate is dropped from the ballot and votes cast for the candidate distributed according to second preferences, and so on until a candidate is elected. The argument is that in ethnically mixed electoral districts, the alternative vote creates the incentive for candidates to appeal across ethnic lines.
• Centripetalists also advocate a presidential form of government, because it would provide another opportunity for vote pooling, if it were properly designed. Two round presidential elections are thought to have this quality. A single president, although from one ethnic group, could legitimately claim to support from more than one ethnic group, and could rise above ethnic politics. By contrast, a power-sharing cabinet does not possess the capacity to transcend ethnicity, since cabinet members are thought to represent specific ethnic groups.
• Finally, centripetalists advocate federalism as a tool to manage ethnic conflict, but of a different form than consociationalists do. They see federalism as a tool to fragment the power of the largest ethnic group and thereby diminish its capacity to capture the entire state. The means for doing so is to ensure that a majority group is divided into a number of different provinces. This will create political cleavages within the group over issues such as the sharing of resource revenues that impair its ability to act collectively. As well, centripetalists promote ethnically heterogeneous provinces to make made allies out of ethnic groups who would otherwise have been competitors, and unite around their shared material interest as residents of the same province.
C. Beyond Integration/Accommodation Dichotomy
The debates between integration and accommodation, and under the accommodationist umbrella between advocates of the consociation and centripetalism, are highly polarizing. Indeed, they can be as polarizing as ethnic conflict itself, not only in the lecture hall, but also in constitutional assemblies and the negotiation table.
I want to come at this debate from a different angle, and suggest three ways of moving beyond it.
First, rather than conceptualizing the choice between integration or accommodation as an all or nothing decision that is made at a single moment for all time, it may be better to understand these models as appropriate for different stages of a constitutional transition. Thus, the question is not integration or accommodation, but rather the sequence between integration and accommodation.
Imagine a constitutional transition as part of a civil war settlement. In a fundamental sense, civil wars are constitutional phenomena.Above all else, the basic mission of constitutionalism is to channel political conflict, disagreements that would otherwise spill into the streets and be settled according to violence, into institutions that operate peacefully according to law and reach decisions that members of a political community accept as authoritative. In civil wars, constitutions do not perform this basic function. Civil wars are started by a conscious decision to step outside the previous constitutional order, and are always justified by a narrative of constitutional failure that explains why that constitutional order was no longer legitimate and worthy of obedience.
In order to settle a civil war through negotiation, a warring party will demand a new constitutional dispensation that rectifies the mistakes of the past. In ethnic civil wars, a warring minority will almost always argue that it was the victim of majoritarian political institutions. It will demand consociational power-sharing as the price for peace, and a majority may have no option. But once the order is operational, and members of ethnic groups develop reciprocal bonds of trust, it might be possible to move to a more integrationist set of political institutions.
The leading example of this kind of sequence is provided by South Africa. During the negotiations for the transition from white to majority rule, the negotiators had conflicting objectives. The white regime, represented by the National Party, would not surrender power without firm guarantees for minority rights in a constitution. The African National Congress argued that any negotiated constitution would be illegitimate unless it was drafted and adopted by a democratically elected constitutional assembly.
The compromise was an interim constitution that came into force in 1993. The interim constitution was a transitional document that would be replaced by a final constitution. It contained a number of safeguards for the white minority to ensure its interests would be protected in the final constitutional document. The final constitution could only be adopted by a 2/3 majority of the Constitutional Assembly. In addition, the final constitution would need to comply with a set of 34 legally binding constitutional principles contained in the interim constitution that protected the vital interests of the white minority. These included requirements for bicameralism, a bill of rights backed up by judicial review, constitutional supremacy, regular elections according to universal suffrage and under proportional representation, super-majority consent for future constitutional amendments, guaranteed minority party participation in the legislative process, federalism, independent institutions, and so on. The Constitutional Court would be required to sign off on compatibility of the constitutional text with these principles.
In addition, there were safeguards for the white minority during the interim period while the final constitution was under negotiation by mandating a power-sharing cabinet. The interim constitution guaranteed a cabinet seat to any party with 5% of the vote, and a deputy president to any party obtaining 25% of the vote. Moreover, this guarantee was to survive the transition to the final constitution for a total of five years.
South Africa provides one illustration how integrationist and accommodationist constitutional arrangements can be put together in a sequence. But it does not exhaust all the possible arrangements, which must ultimately be driven by national context.
Second, proponents of integration and accommodation suggest that each amounts to a comprehensive and mutually exclusive constitutional strategy. However, there is a gap between theory and practice. Many actual constitutions often contain a mixture of integrationist and accomodationist elements.
Consider the example of my own country’s constitution, Canada. Canada is often offered as the paradigmatic instance of a multinational federation that accommodates its large territorially concentrated French-speaking minority. The Canadian Constitution creates a federation in which the boundaries of the province of Quebec are drawn so that French-speakers constitute a majority and are not outvoted by Canada’s English-speaking majority. In addition, Quebec has jurisdiction over policy areas to ensure the survival of a French-speaking society, including language and education. Quebec has the power to set its own official language, and has taken measures to ensure that French is the common language of economic and political life.
But many important areas of interest to the citizens of Quebec lie under federal jurisdiction, such as immigration policy, income taxation, foreign policy, and defence. Citizens of Quebec must participate in the institutions of the federal state in order to shape important decisions that affect them. In addition, the institutions of shared rule at the federal level are somewhat accommodationist, but mostly integrationist. The major accommodationist element in federal institutions is the official language status of French. On the integrationist side of the ledger, the House of Commons is directly elected; the House of Commons and Senate pass bills by a simple majority vote, with no formal role for the Quebec government, or Quebec members of Parliament or Quebec ministers in the federal legislative process; the constitutional conventions of responsible government for a Westminster democracy apply without any modification for the fact that Canada is a multinational state.
Why are real constitutions hybrids, like the Canadian Constitution? Constitutions are the result of messy political compromises. But there is logic to these hybrid arrangements. Multinational federalism has as its goal the preservation of the territorial integrity and political unity of the state. But it can perversely fuel secession. Integrationist elements offset this danger, by binding the subunit to the state, for example, by requiring its citizens to participate in shared institutions on a footing of equality. Quebecers participate in federal institutions not as Quebecers, but as Canadians. Indeed, Quebec’s elites have long reasoned that even in a decentralized federal state such as Canada, they ignore the structure of the center at their peril, and have always sought to ensure that Quebec’s presence in federal institutions is not diluted.
Another hybrid dimension to the Canadian Constitution is that Quebec’s extensive jurisdiction is subject to a national bill of rights, the Canadian Charter of Rights and Freedoms. One of the arguments frequently advanced against the accommodation of minority nationalism through federalism is that it may lead to the creation of local tyrannies. Ethnic minorities who constitute a provincial majority might view the province as belonging to them rather than to all the province’s residents on a footing of equality. A possible result might be a “sons of the soil” politics that legitimizes discrimination against internal minorities in the framing of public policy, the delivery of public services, contracting, and public sector employment – especially if the internal minorities are members of the national majority. Through its provisions for equality rights and interprovincial mobility rights, the Canadian Charter rules out policies that openly discriminate on the basis of ethnic identity or against recent migrants from other provinces.
Tl\ird, although integration and accommodation are often diametrically opposed, they have a common enemy: democratic authoritarianism. Democratic authoritarianism develops when a directly elected President harnesses legislative majorities to expand executive power through constitutional amendment. For example, these amendments may remove Presidential term limits. They may also erode the independence of institutions that are designed to counterbalance the political executive, such as an independent judiciary, a non-partisan police force, and a professional civil service. Coupled with a compliant government caucus, held together by patronage and fear, the executive is subject to few checks and balances.
In a plural society, democratic authoritarianism poses a threat to the constitutional agendas of both accommodationists and integrationists. Democratic authoritarianism, by its very nature, does not share executive power. Cabinets that purport to be coalitions of different parties representing different ethnic groups mask the reality that ethnic minorities do not participate in the exercise of power under authoritarian rule. Rather, they are co-opted by authoritarians to lend them legitimacy. Democratic authoritarianism is also hostile to the political competition favoured by integrationists. Democratic authoritarians will seek to ensure that their political parties achieve dominant status by manipulating electoral rules, interfering with fair elections, and deploying the largesse of the state to underwrite its electoral support. In a plural society, democratic authoritarianism blends partisan abuse with ethnic majoritarianism.
It is politically vital for proponents of integration and accommodation to find points of agreement on specific constitutional arrangements to make a common front against democratic authoritarianism, even if they continue to disagree at a more fundamental level on a range of other issues. I want to conclude by suggesting that devolution can serve as a point of strategic overlap between the integrationist and accommodationist constitutional agendas. I have already addressed the idea that devolution can accommodate a territorially concentrated ethnic minority. But it may also enhance political competition at the national level and may accordingly serve as a check on the rise of, and may precipitate the decline of, dominant political parties.
The reason is that devolution multiplies the opportunities for electoral choice and political competition in two ways. First, devolution increases the number of governments that must be democratically elected. Second, it creates different political majorities empowered to elect different governments-a national majority, and two or more provincial majorities. The proliferation of opportunities to wield power creates the space for political parties that lose at the national level and are relegated to opposition status nationally to win at the provincial level through the support of a different political majority.
Moreover, provincial governments provide important political resources to parties that strengthen their ability to compete at the national level. The possibility of wielding power enhances the ability of parties to recruit and train political elites. The expertise developed from mass political mobilization at the provincial level during provincial elections can be transferred to national elections. Governing at the provincial level provides parties with the advantages of incumbency, such as greater public profile and the ability to shape public policy to enhance their base of political support.
Perhaps the most dramatic example of the effect of devolution to increase political competition at the national level and precipitate the decline of a dominant political party is India. The Congress Party dominated Indian politics for the two decades after independence, winning continuous majorities at the national level as well as majorities in most states. The success of the Congress Party turned on its internal structure, which in turn was a function of India’s electoral system. India’s system of single- member constituencies elected by plurality placed great importance on state-level political organizations. As a consequence, the Congress Party was heavily reliant on state-level political elites.
In the 1967 elections, Congress faced a successful challenge from new regional parties at the state level, and lost power in eight states. Victories by regional parties deprived the Congress Party of the power of patronage at the state level, which limited its ability to mobilize support in the national elections. Ultimately, this culminated in outright losses by Congress in 1977,1989, and 1996 in national elections. The Congress Party has never again obtained an outright majority of seats.
India illustrates that devolution can further multiple constitutional agendas. India’s federal structure has served as a vehicle to both accommodate minorities, and to promote political pluralism. The broader lesson is this. Accommodationists and integrationists should search for other points of contingent agreement around which to build an alliance against democratic authoritarianism. If they do not, the choice will not be between integration and accommodation. There won’t be any choice at all.
